Precipitation remains a challenging process for weather and climate models to simulate accurately, while changes in its spatial-temporal variability yield critical scientific and societal implications in a warming climate. To better address global and regional precipitation variability, we aim to systematically quantify the attribution of precipitation and its extremes to a set of weather systems objectively identified by individual algorithms, including atmospheric rivers (AR; TECA Bayesian AR Detector (TECA-BARD)), mesoscale convective systems (MCS; FLEXTRKR), low-pressure systems (LPS; TempestExtremes) and fronts (FT; Sansom-Catto algorithm). By merging Global Precipitation Mission precipitation and these identified atmospheric features using ERA5 reanalysis, a long-term, global dataset of feature-associated precipitation is generated to robustly break down rainfall contributions from individual feature types and their co-occurrence, considered crucial for precipitation extremes.
Precipitation associations with individual feature types are in accord with prior studies, but precipitation attribution to multiple features, i.e. co-occurrence, is widespread. For instance, AR-FT co-occurrence explains at least 75% of precipitation extremes exceeding the 99th percentile over most North American Pacific Northwest regions, followed by the individual contribution from FTs. MCSs dominate extremes over tropical oceans, as expected from previous literature. Extremes over the Amazon exhibit clear contrast between coastal and inland areas, related to MCSs and isolated deep convection, respectively. In addition, the MCS-FT co-occurrence contributes to most extremes over the South Atlantic Convergence Zone in South America. Feature-precipitation relationships to key environment variables are also investigated to develop process-oriented diagnostics. Overall, preliminary work demonstrates the value of coordinated efforts in quantifying precipitation attribution and the potential for model assessments of precipitation association with underlying phenomena.